The TUM coordinates the InShaPe project consortium, which comprises 11 members from France, Israel, Italy, the Netherlands, Sweden, Slovenia, Spain, and Germany. The project's goal is to establish a novel first-time-right laser-based powder bed fusion of metals (PBF-LB/M) process that enables green AM with much higher build rates at a lower cost per part, less energy consumption and less scrap. Two major technical innovations drive this approach for sustainable AM: an artificial intelligence (AI)-based beam shaping module that can generate arbitrary beam shapes and a multispectral in-line monitoring and control system for understanding, improving, and controlling the manufacturing process with different beam shapes.
The beam shape and the processing strategy heavily influence the process stability, the melt pool geometry, and the resulting microstructure. To investigate the different influences, an EOS M290 3D printer with a beam shaping module and a multispectral camera system was installed at the TUM to conduct research within the InShaPe project. With this new system, running a PBF-LB/M process with arbitrary beam shapes and monitoring the resulting melt pool to gain insights and draw conclusions is possible. Besides the standard Gauss shape, different ring-core shapes, a C-shape, and a computer-simulated "Chair"-Shape were already tested on the machine and showed the first promising results:
- Improvement of process stability
- Improvement of part quality
- Tailoring of the microstructure possible
- Widening of the process window
Richard's task is creating an AI-based "recipe book" framework that uses all the gathered process data to explore connections between the part geometry, beam shape, process parameter, laser path, and resulting part properties such as the microstructure, mechanical part properties, or surface quality. The following steps will be a prototype framework, and more experiments with different beam shapes, and various process parameters.n und Prozessparametern.